1. Field of the Invention
The present invention relates to a semiconductor memory device and method of fabricating same. More particularly, the present invention relates to capacitor fabrication techniques applicable to dynamic random access memories ("DRAMs") capable of achieving an improved degree of integration and a lower number of defects within the DRAM.
2. State of the Art
A widely-utilized DRAM (Dynamic Random Access Memory) manufacturing process utilizes CMOS (Complementary Metal Oxide Semiconductor) technology to produce DRAM circuits which comprise an array of unit memory cells, each including one capacitor and one transistor, such as a field effect transistor ("FET"). In the most common circuit designs, one side of the transistor is connected to external circuit lines called the bit line and the word line, and the other side of the capacitor is connected to a reference voltage that is typically one-half the internal circuit voltage. In such memory cells, an electrical signal charge is stored in a storage node of the capacitor connected to the transistor which charges and discharges circuit lines of the capacitor.
Higher performance, lower cost, increased miniaturization of components, and greater packaging density of integrated circuits are ongoing goals of the computer industry. The advantages of increased miniaturization of components include: reduced-bulk electronic equipment, improved reliability by reducing the number of solder or plug connections, lower assembly and packaging costs, and improved circuit performance. In pursuit of increased miniaturization, DRAM chips have been continually redesigned to achieve ever-higher degrees of integration which has reduced the size of the DRAM. However, as the dimensions of the DRAM are reduced, the occupied area of each unit memory cell of the DRAM must be reduced. This reduction in occupied area necessarily results in a reduction of the dimensions of the capacitor, which, in turn, makes it difficult to ensure required storage capacitance for transmitting a desired signal without malfunction. However, the ability to densely pack the unit memory cells while maintaining required capacitance levels is a crucial requirement of semiconductor manufacturing technologies if future generations of DRAM devices are to be successfully manufactured.
In order to minimize such a decrease in storage capacitance caused by the reduced occupied area of the capacitor, the capacitor should have a relatively large surface area within the limited region defined on a semiconductor substrate. The drive to produce smaller DRAM circuits has given rise to a great deal of capacitor development. However, for reasons of available capacitance, reliability, and ease of fabrication, most capacitors are stacked capacitors in which the capacitor covers nearly the entire area of a cell and in which vertical portions of the capacitor contribute significantly to the total charge storage capacity. In such designs, the side of the capacitor connected to the transistor is generally called the "storage node" or "storage poly" since the material out of which it is formed is doped polysilicon, while the polysilicon layer defining the side of the capacitor connected to the reference voltage mentioned above is called the "cell poly."
An article by J. H. Ahn et al., entitled "Micro Villus Patterning (MVP) Technology for 256 Mb DRAM Stack Cell," 1992 IEEE, 1992 Symposium on VLSI Technology Digest of Technical Papers, pp. 12-13, hereby incorporated herein by reference, discusses the use of MVP (Micro Villus Patterning) technology for forming a high surface area capacitor. FIGS. 25-28 illustrate cross-sectional views of this technique. FIG. 25 shows a memory cell structure comprising a substrate 202 which has been oxidized to form thick field oxide areas 204 with transistor gate members 206 disposed on the surface of the substrate 202. A barrier layer 208 is disposed over the transistor gate members 206, substrate 202, and field oxide areas 204, and a silicon nitride layer 210 is disposed over the barrier layer 208. A storage poly 212 is disposed on the silicon nitride layer 210 and extends through the silicon nitride layer 210 and the barrier layer 208 and between two transistor gate members 206 to contact the substrate 202. A layer of silicon dioxide 214 is disposed over the storage poly 212.
As shown in FIG. 26, an HSG (HemiSpherical-Grain) polysilicon layer 216 is grown on the exposed surfaces of the silicon nitride layer 210, the storage poly 212, and the silicon dioxide layer 214. The structure is then etched using the HSG polysilicon layer 216 as a mask which results in very thin, closely spaced micro villus bars or pins 218, as shown in FIG. 27. The silicon dioxide layer 214 and the silicon nitride layer 210 are then stripped to form the structure shown in FIG. 28. A finalized capacitor would be formed by further processing steps including depositing a dielectric layer on the etched storage poly and depositing a cell poly on the dielectric layer.
Although the MVP technique greatly increases the surface area of the storage poly, a drawback of using the MVP technique is that it can result in splintering problems (or slivers) in the storage node cell poly. As illustrated in FIG. 29, the micro villus bars/pins 218, formed in the method shown in FIGS. 25-28, are thin and fragile such that they are susceptible to splintering that may result in one or more of the micro villus bars/pins (such as pin 220) falling over and shorting to an adjacent storage poly 222, which would render the adjacent storage cells shorted and unusable.
In a 64 M DRAM, for example, even if there was only one out of 100,000 cells that had a failure due to a splintered macro villus bar/pin shorting with an adjacent storage cell, it would result in 640 failures or shorts in the DRAM. Generally, there are a limited number of redundant memory cells (usually less than 640 in a 64 M DRAM) within a DRAM which are available for use in place of the shorted memory cell. Thus, if the number of failures exceeds the number of redundant memory cells within the DRAM, the DRAM would have to be scrapped.
Therefore, it would be desirable to increase storage cell capacitance by using a technology such as MVP while eliminating polysilicon storage node splintering problems.